[Chimera-users] surface failures

[Elaine Meng]

On Apr 22, 2009, at 1:26 AM, Cédric Govaerts wrote:

> Hi Elaine,
> Here is my problem, seems like I cannot compute these surfaces on  
> some PDBs
> and receive the following message in the Reply Log:
>
> C:\Program Files\Chimera\bin\mscalc.exe 1.400000 2.000000 1
> Calculation of some surface components failed (code 5).
> Falling back to single-component calculation.
> C:\Program Files\Chimera\bin\mscalc.exe 1.400000 2.000000 0
> Surface calculation failed, mscalc returned code 5
>
> I attach the PDB so you can try it for yourself.

Hi Cedric,
This is an uber-common problem in Chimera due to the numerical failure  
of the embedded MSMS code on some structures.  There are many  
workarounds that are slight changes to the surfacing parameters or VDW  
radii, but it is an art rather than science to find them -- just  
requires trial and error and typically is different on different  
machines.  For example, on both my mac 10.4 laptop and mac 10.5  
desktop machines, I just opened your structure and showed the surface  
without any problems or doing anything special.

Because of the MSMS problems, we have been working for many months on  
code to replace it.  We are hoping to have our own surfacing code  
ready for the next production release.

In the meanwhile, I will briefly summarize some things to try (but be  
aware they can affect your surface area results):

- if your structure has multiple chains and you don't mind having a  
separate surface for each chain, use the "split" command to split it  
into separate models before surfacing

- before creating the surface, adjust related parameters in the New  
Surfaces category of Preferences (under Favorites).  For example, you  
can adjust vertex density and/or probe radius. if you only care about  
the main outer surface and not interior bubbles you can turn off "show  
disjoint".  Actually you might prefer not to have interior bubbles  
included in your results.  Sometimes the components that fail are just  
those interior bubbles, and "single-component calculation" as  
mentioned in your message of the main outer surface still works.   
However, in your case both failed.

- slightly change VDW radii, e.g. with command "vdwdefine +.05" (you  
can put them back to default with command "~vdwdefine")

Here are several previous posts on this issue:
<http://www.cgl.ucsf.edu/pipermail/chimera-users/2008-September/003096.html 
 >
<http://www.cgl.ucsf.edu/pipermail/chimera-users/2008-March/002417.html>
<http://www.cgl.ucsf.edu/pipermail/chimera-users/2008-August/ 
002954.html>
>
>> ("component") are reported in the Reply Log.  At the same time, it
>> creates atom and residue attributes named areaSES and areaSAS.
>
> Hum, anyway I can retrieve such per residue information ?
> Maybe in a text file or something ?
> Just to make sure, I can compute the SAS for each atoms based on the
> complete structure (including sidechains) and then obtain the  
> individual
> values for each atoms (say backbone nitrogens in this case.)

You can save the attribute values (atom or residue attribute named  
areaSAS or areaSES) to text files using the Render by Attribute  
dialog's File menu.  I will mail you the text file for atom attribute  
areaSAS for your structure in a separate message.

I hope this helps,
Elaine
-----
Elaine C. Meng, Ph.D.                          meng at cgl.ucsf.edu
UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab
Department of Pharmaceutical Chemistry
University of California, San Francisco
                      http://www.cgl.ucsf.edu/home/meng/index.html